DESCRIPTION (As Adapted from the Investigator's Abstract): Homeless families headed by women constitute the fastest growing subgroup of the homeless population. While many issues of homelessness can be traced to broad economic and social factors, preliminary studies indicate a number of personal factors can act as barriers to seeking health care. Therefore, the specific aim of this pilot study of a modified Social Support Intervention are to: (1) increase homeless women's health maintenance behaviors; (2) increase homeless women's general health; (3) decrease homeless women's use of emergency room visits; and (4) explore factors that influence homeless women's health-related experiences. To achieve these aims the proposed pilot study will use a quasi-experimental design. To determine if a modified Social Support Intervention improves health outcome among homeless women living in a transitional shelter, and experimental group of 24-30 women will receive the intervention and the health outcomes of this group will be compared to the health outcomes of 24-30 control group participants. The modified Social Support Intervention will consist of four social support sessions for 8 weeks followed by monthly individual sessions and group sessions at 6 months and at 12 months. The health outcomes will be measured by the Health Maintenance Behavior Check List, the Short Form Health Survey, SF-36, and a question about emergency room use. To obtain a profile of participants, measurement will include the Sociodemographic Scale, the Addiction Severity Index and the Symptom Checklist 90-r, and the Abuse Assessment Screen. To explore factors that influence health, participants will complete the modified Prenatal Psychosocial Profile which measures: stress, social support and self-esteem. The ethnically diverse sample of homeless women will be recruited from a transitional shelter in southern California. Data will be collected individually from the women to allow for confidentiality. Quantitative data will be analyzed using descriptive and inferential statistics. Qualitative data from interviews will be described verbatim and analyzed using grounded theory and dimensional analysis. The long-range goal of this proposed pilot study is to provide a basis for an intervention program for homeless women that will improve their health outcomes.